Abstract / Summary
Abstract Background In adolescent and young adult (AYA) women, high-risk human papillomavirus (HR-HPV) infection is often transient, so direct colposcopy referral of all HR-HPV-positive women may create avoidable burden. In Chinese hospital-based screening, HPV DNA genotyping and HPV E6/E7 mRNA assays coexist, but AYA-specific triage evidence is limited. We evaluated age-specific screening findings and modelled referral-detection trade-offs, treating platform comparisons as ecological. Methods We analysed women undergoing opportunistic cervical screening at a tertiary hospital in Henan, China, in 2023–2024. Women were classified as AYA (15–39 years) or non-AYA ( > = 40 years); 32,737 with valid cytology formed the analytic cohort. HR-HPV testing used either the Kaineng 23-genotype DNA assay or the Hologic Aptima E6/E7 mRNA assay. The primary endpoint was histologically confirmed cervical intraepithelial neoplasia grade 2 or worse (CIN2+); the secondary endpoint was grade 3 or worse (CIN3+). We estimated marker positivity, triage referral, histology-verified CIN2 + capture, platform concordance, and inverse-probability-weighting (IPW)-corrected diagnostic accuracy. Results HR-HPV positivity was lower in AYA than non-AYA women on both platforms (DNA 15.7% vs 18.7%; mRNA 8.3% vs 12.8%; both p < 0.001). Among verified women, CIN2 + prevalence was similar (8.8% vs 8.5%), whereas CIN3 + was lower in AYA women (1.7% vs 3.0%; odds ratio 0.56, 95% confidence interval 0.35–0.89). In AYA co-tested women, cytology triage of mRNA-positive women referred 4.0% and captured 73.4% of verified CIN2+; the DNA pathway referred 6.1% and captured 68.8%. Genotype-only triage minimized referrals but missed many verified CIN2 + cases. Among 864 dual-tested women, platform agreement was moderate (kappa = 0.42). Naive HR-HPV sensitivity of 82–87% fell to 43–51% after IPW correction. Conclusions AYA women did not have a higher verified CIN2 + burden and had fewer CIN3 + lesions than older women. Cytology-directed triage appeared to reduce colposcopy compared with direct referral. The apparent mRNA pathway advantage should be interpreted cautiously because platform assignment was non-random and mostly non-overlapping. Paired prospective studies with complete verification and follow-up are needed.